The Liquid Metal Cooling Process (LMC) is a well known variation of the investment casting directional solidification process in which a shell mold which forms a cavity resembling the final part to be manufactured is filled with a liquid alloy, and lowered from a hot zone into a cold zone. The hot zone is usually maintained at above the temperature of solidification of the alloy being used and the cold zone is well below this temperature. As the mold is drawn at a controlled rate into the cold zone, solidification occurs directionally beginning with the first portion of the mold to enter the cold zone. The mold may be supported on a chill plate used for a structural base and also to rapidly remove heat from the bottom-most portions of the casting. The typical Bridgman process (see, for example, U.S. Pat. No. 1,793,672) uses a water-cooled copper or steel lined vacuum chamber as a means of cooling the shell mold by radiation. There is a baffle made of an insulating material between the cold zone and hot zone which prevents excessive heat flow therebetween. The baffle is shaped so as to let the shell mold pass through it with only a minimum amount of clearance space between the mold and the baffle. This further minimizes heat flow from the hot zone to the cold zone, maintains the solidification front of the material in the shell mold at a level close to that of the baffle, which is desirable, and maximizes the thermal gradient across the solidification front.
The LMC process (see, for example, U.S. Pat. No. 3,763,926) uses a liquid medium (usually aluminum or tin) kept at a controlled temperature as the cold zone. While U.S. Pat. No. 3,763,926 discloses a stationary baffle similar to that used in the Bridgman process, a variation of this, using a baffle which floats on the surface of the liquid coolant, is disclosed in U.S. Pat. No. 4,108,236. Those familiar with the art of directional solidification understand that the rate of heat transfer achievable at the surface of the shell mold is much greater when using liquid metal as a coolant than simply relying on radiation cooling in vacuum. It can be shown by calculations that the effect of LMC is 70 to 100% greater heat transfer away from the shell mold surface compared to pure radiation heat transfer.
The Bridgman and LMC directional solidification processes have been used for the industrial production of columnar grained and single crystal articles destined for high temperature service and composed of Ni-based superalloys. Those familiar with the solidification of Ni-based superalloys understand the great advantages of the higher gradients across the solidification front achieved during the LMC process; as outlined by Giamei and Tschinkel in Metallurgical Transactions A, Volume 7A, September 1976, pp. 1427-1434, these lead to finer dendrite arm spacings, smaller pores, the suppression of freckle defects, and smaller diffusion distances required during solution heat treatment. The effect of high gradient solidification on the lifetime of single crystal Ni-based superalloys is well known. For example, commercially available CMSX-2 exhibits approximately a doubling in Low Cycle Fatigue life and an order of magnitude improvement in creep rupture life at 870.degree. C. when solidified with a high gradient process compared to a low gradient process (see page 1005, "Directionally Solidified and Single Crystal Superalloys"--The Metals Handbook, Tenth Edition, Volume 1, ASM International 1990).
However, it is also well known in the industry that there are relatively few LMC furnaces in production compared to the very numerous Bridgman furnaces. One reason for this is the very high cost of the LMC furnace, i.e., maintaining a bath of liquid aluminum or tin at a controlled temperature is very complicated and costly. The high cost offsets the improvements in material properties achieved by using the process in most cases. One way to improve this situation would be to increase the production rate at which components are cast using the LMC process. As outlined by Giarnei and Tschinkel, as the rate of withdrawal of the shell mold from the hot zone into the liquid coolant increases, so does the curvature of the solidification front. This is highly undesirable (leading to freckle defects and other problems) and so places an upper limit on the rate of withdrawal and hence the production rate of LMC cast components. What is needed is a means of avoiding this solidification front curvature at higher withdrawal rates, and generally increasing the amount of heat transfer in the process, in order to increase production rates using LMC and hence lower the cost of the process and further improve material properties.